Dependency-Aware Adaptive Test Healing with Graph Neural Networks

Publication at a glance

Paper: Enhancing Adaptive Test Healing with Graph Neural Networks for Dependency-Aware Decision Making
Authors: Nariman Mani, Salma Attaranasl
Published in: IEEE AITest 2025
Persistent record: 10.1109/AITest66680.2025.00023
Google Scholar: View the publication record

The research problem

A broken test rarely exists in isolation. It depends on application components, test fixtures, APIs, data, and other execution steps. A healing action based only on the final error can repair the wrong location or mask a real regression.

The paper's central contribution

The paper brings graph neural networks into adaptive test healing so decisions incorporate dependency structure. Graph representations let a model reason about neighborhoods and propagation paths rather than treating each failure as a flat feature vector.

This distinction matters for both researchers and practitioners: the contribution is a way to reason about the complete system and its decision boundaries, rather than a claim that adding one AI model solves the underlying engineering problem.

How the approach works

A dependency graph represents relationships among tests and relevant software entities.
Graph neural message passing builds context from connected nodes and their features.
A decision layer uses that structural representation to select or rank healing responses.

Together, these elements make the approach easier to study, reproduce, and extend. They also provide clear terms for literature searches and comparisons with related work.

Why this publication matters

The contribution is relevant to self-healing test automation, flaky-test diagnosis, graph-based software analytics, and intelligent CI/CD. It supplies a bridge between program structure and learning-based recovery decisions, especially in systems where a local symptom can have a remote cause.

For practitioners, the larger lesson is to measure the system behavior that matters—not only a model-level score. Reliability, privacy, change over time, human review, and downstream consequences belong in the evaluation.

Research questions this work can support

This publication is relevant when investigating questions such as:

How should the core problem be represented so an AI-assisted system retains the right context?
Which technical and organizational signals provide early, actionable evidence?
How can automation improve adaptability without concealing failures or weakening safeguards?
What evaluation design captures effectiveness, safety, privacy, and long-term change?
Where should a system defer to a developer, architect, coach, clinician, or participant?

Scope and responsible interpretation

A graph is only as useful as its nodes, edges, and freshness. Dynamic dependencies, incomplete traces, concept drift, and unseen failure types can reduce reliability. Safe adoption needs confidence thresholds, auditable recommendations, regression checks, and an option to decline automated healing.

These boundaries are opportunities for replication, comparison, and follow-on research. Readers should consult the paper itself for its precise method, datasets, experimental setup, and reported results.

How to cite this paper

Mani, N., & Attaranasl, S. (2025). Enhancing Adaptive Test Healing with Graph Neural Networks for Dependency-Aware Decision Making. IEEE AITest 2025. https://doi.org/10.1109/AITest66680.2025.00023

Use the 10.1109/AITest66680.2025.00023 publication page to verify the latest bibliographic metadata before submission. You can also find the work through its Google Scholar record. If your research builds on the architecture, method, evaluation, or research framing described here, please cite the original publication rather than this explanatory post.

Related research on this weblog

Browse the complete Nariman Mani publications list.
Explore GraphQL-aware healing with multi-signal learning.
For adjacent work, read adaptive test healing with LLMs and reinforcement learning and privacy-by-design adaptive group assignment.